It is more to do with volumes of product, hardware is costly and the margins are thin if targeting less than 1000 per year. The problem is the product is not versatile to be adopted in a more general sense. The Engineering accomplishment of the company are great, but it does not have a broad enough market and hence reflected in the cost of the product. Similarly, you can compare that to technical books and why they cost so much as the expected sales volumes are not high enough unlike a fictional character book, prime example Harry Potter, it also became a big screen hit with fans.

Hiring cheap human labor in the US costs (for minimum wage) $36 to $40k per year.

So according to [1] spot costs 1 year of a decent-wage employee or 2 years of a minimum wage employee. And of course that's disregarding that spot will happily pull nighttime duty, work on weekends, Christmas, and will happily walk into a chemical plant risking his life for you ...

[1] https://beebole.com/blog/how-to-calculate-the-real-cost-of-a...

The problem with Boston Dynamic's robots imo isn't the cost versus a human but the cost versus a simpler and cheaper yet just as capable (if not more capable) wheeled robot.

I have a very hard time coming up with applications where the walking capability is useful. On the other hand the space wasted to put all this technology could be used for more batteries, a sturdier body to carry heaviers weights, etc.

The advantage of legs is for use in environments set up for humans with working legs. Things like a normal flight of stairs, walking over a moderate sized gap in the floor, etc.

Those sorts of things can be a real pain for wheeled robots (things like stair climbing wheel assemblies, and just larger wheels can overcome some of these, but have their own downsides), while spot has no issues handling them.

If you can design the environment to fit the needs of the robots, then sure wheeled or even on-rails will likely be a better choice (less expensive, greater load, etc) But for retrofitting robots into existing environments without totally reconfiguring them legs can be rather useful.

The question is are there enough such circumstances that Boston Dynamics can actually make reasonable amounts of money? I'm not sure. It is possible there are some fields with a lot of killer applications for such robots, but few spring to mind.

But that human has rights and can't work 24 hours a day...

neither can the robot, I think it lasts 90 minutes or so. And employers don't mind human rights if it's still cheaper than the machine

The point of BD is to build robots that can move like humans in conditions that would kill a human.

Humans usually charge quite a lot for suicide missions.